A Beginner’s Guide to NHL Overlays and Underlays
Why the odds feel off
You’re staring at the board, the numbers look clean, but something smells wrong. The house edge is hidden in plain sight, and you’re about to get burned if you don’t spot the twist. Here’s the deal: overlays and underlays are the secret sauce that separates the casual fan from the profit machine.
What an overlay actually is
Think of an overlay like a safety net under a high wire. The bookmaker’s line says “Toronto will win at -150,” which translates to a 60% implied probability. If your own model says the Puck Club has a 68% chance, you’ve uncovered an eight‑point overlay. In plain English, the odds are generous enough that a smart bet should actually be a win.
Underlays — the opposite trap
Now flip the coin. An underlay is a thin ice sheet you don’t want to skate on. The bookies might list the Colorado Avalanche at +120, implying a 45% chance, but your analysis says they only have a 30% shot. That’s a 15‑point underlay, a red flag that the market is overvaluing the team. Bet the underlay only if you’re hunting a hedge or a long‑term bankroll builder.
How to calculate implied probability
Take the money line, drop the sign, then divide 100 by (100 + line) for positive odds or (100 – line) for negatives. It’s math, not magic. Once you have that percentage, line it up against your own projection. The gap is the overlay or underlay. No fancier formulas, just raw numbers.
Finding real overlays on the web
Sites like hockey-bets.com publish constantly updated odds tables. Scroll, compare, subtract. The moment you see a line that consistently sits below your model’s probability, you’ve struck gold. Stop chasing the hype; trust the data.
Risk management in practice
Don’t go full‑tilt on every overlay. Use a Kelly‑type fraction: stake = edge / odds. If you have a 5% edge on a -200 line, that’s a 2.5% bankroll bet. Keep the unit size small, keep the variance low. Discipline beats excitement every time.
Final actionable tip
Pick one game, compute the implied probability, compare it to your model, and place a bet only if the edge exceeds 3%. That’s it.
← Back to Blog